Assumptions and limitations TOOL METHODOLOGY

Water Risk Valuation Tool 14  WRI water scarcity projections for 2020 and 2030.  Shadow water prices TEVs 4. Finally, to simplify the exercise, we have relied on linear projections based on historical growth, and thus we have also limited the number of years for which we project earnings and share prices. Similar to our gold price assumptions, these forecasts will need more rigorous modelling in future iterations in order to expand the scope to time horizons more often used by analysts i.e., closer to a decadal timeframe. To help address these limitations, the WRVT enables users to reach beyond its boundaries to maximize flexibility through its override fields. An analyst with access to more granular information or a different view on scenarios can fine-tune many of the key inputs to the model by overwriting the model’s defaults. Water Risk Valuation Tool 15

4. APPLICATIONS AND EXAMPLES

This section outlines some case studies to illustrate how analytics and insights from applying the WRVT can be used to assess investment risk and inform corporate engagement. We present four use cases that illustrate in broad strokes how different outputs of the tool can be utilized by analysts. Case Study 1: Identify macro-level company exposure to water risk The tool provides a unique macro-level view of water stress exposure based on WRI data by aggregating asset-level data on water stress severity and the proportion of total company output located in areas of water stress. This can inform assessments of how resilient a company’s overall operations are in the face of current and future water constraints, as demonstrated in Figure 3 below. For instance, the tool shows that 80 of Capstone’s copper production is located at the water-stressed Pinto Valley mine in the United States and Cozamin mine in Mexico. By 2020:  61 of the company’s production may experience an increase in water stress by 1.4 times baseline  37 of total production output will hold steady at 2010 levels of water stress Figure 3. Capstone: Change in of production exposure to water stress between 2010 and 2020 Case Study 2: Model potential unanticipated cash flow fluctuations A mining analyst could use the WRVT to model potential unanticipated cash flow fluctuations from potential production loss andor increased costs to maintain production under conditions of water stress. Figure 4. The WRVT enables users to easily model and observe the impacts of water stress on mining equity valuations Water Risk Valuation Tool 16 For instance, using the default settings on the WRVT demonstrates that a water supply gap could theoretically have an impact on Antofagasta Figure 4:  The model shows a 40 difference in free cash flow between the DCF outputs that incorporate and exclude water risk considerations model estimate of 3,817 million considering water risk vs. 6,352 million business as usual in 2021.  This in turn would impact Antofagasta’s total equity value and result in a lower projected share price. It is important to note that these figures are not to be interpreted in absolute terms as price targets, but rather a demonstration exercise to observe the relative differences between model results for business as usual “Model” in Figure 4 and water risk considered “wStranding” in Figure 4. Case Study 3: Pricing risk using a customizable framework and shadow prices For the investor already factoring in ESG issues, the tool can complement existing assessments of water use by miningextractives companies. The tool could be a potentially useful identifier of companies in a portfolio with the most or least water risk and help pinpoint underlying factors to more accurately price water risk. For example, the analyst can, in partnership with the client and the credit team, customize the dynamic framework to price risk and account for mitigation efforts:  Combine shadow prices with water use, production and reserves data, to identify stress points and evaluate potential limits to expanding ore extraction at specific mines  Use data provided by the company itself to override estimated information or assumptions in the model to take into account actual water-related expenditure and risk management  Incorporate forecasted expenditure on water rights, water supplies, and water treatment and recovery activities to adjust potential future capex and opex costs. This can provide a more refined, quantitative overview of the risk, which, if significant, can also inform credit conditions. For instance, integrating shadow water prices for 2020 into the model see Figure 5 results in a potential 10 or higher fall in the share prices of several companies analyzed. The WRVT results can be used to guide further qualitative research into how these risks might play out for the companies and sites that are most exposed, by evaluating management responses and policy environments at production sites most exposed to water stress. If an asset manager believes water governance and infrastructure will address water scarcity at a specific location for the lifetime of the mine, the analyst may lower estimates of production losses modelled in the WRVT. Figure 5. TEV costs for a company from 2010-2021